Intelligent robot dynamic cooperation method based on multilayer network and difference accumulation strategy
Through a dynamic collaboration method combining multi-layer networks and differential cumulative strategies in intelligent robots, the problem of inefficiency in existing technologies in complex environments is solved, and higher cooperation rates and adaptability are achieved, suitable for port logistics and unmanned Real-time decision-making requirements in the field of machine assembly.
Patent Information
- Application Number
- CN202411005133.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-12
AI Technical Summary
The existing intelligent robot control methods are inefficient in complex environments, especially in the fields of port logistics and drone shape assembly, where there is less combination of multi-layer networks and evolutionary game theory, resulting in algorithms relying on external parameter transmission in improving group cooperation rate and The performance is average and it is difficult to adapt to environmental changes.
Using a dynamic collaboration method of intelligent robots based on multi-layer network and differential accumulation strategies, optimize behavioral strategies through real-time feedback and continuous evolution algorithms, build multi-layer network models and game matrices, and adjust them using Barabási-Albert Model and Xalvi-Brunet and Sokolov algorithms. The network, combining prisoner's dilemma and public goods game models, performs differential cumulative strategies to improve cooperation rates.
It improves the cooperation rate and adaptability of intelligent robots in complex environments, enhances the robustness and accuracy of the system, reduces experimental time, and is suitable for real-time decision-making needs in the fields of port logistics and drone assembly.
Smart Images

Figure CN118940790A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent robot technology, and in particular to an intelligent robot dynamic collaboration method based on a multi-layer network and a difference accumulation strategy. Background Art
[0002] At present, intelligent robots can be applied in many fields, such as port logistics and drone assembly. In the rapidly developing global trade, ports, as key nodes for cargo transportation, have extremely high requirements for efficiency and reliability. Traditional port handling work relies on a large amount of manpower and simple machinery, facing problems such as low efficiency, high safety risks and high costs. Although the introduction of intelligent robots has alleviated the above problems to a certain extent. However, the terrain and weather conditions in ports are complex and are not suitable for controlling robots by wire. Drone shape assembly is an emerging field, which is currently mainly concentrated in the research and development stage. This technology allows drones to change their shape during flight to adapt to different environments and mission requirements. However, this field faces several challenges. First, shape changes increase the complexity of drone systems and require complex control algorithms. In addition, the application of drone shape assembly is also subject to technical limitations, especially when a global task allocation algorithm is used, which places extremely high requirements on communication and information transmission between drones.
[0003] Among the existing methods for controlling robots, the study of complex networks and multi-layer networks has developed significantly in recent years, especially in the application of evolutionary game theory. The introduction of a large number of new network models not only provides a framework for understanding a variety of real-world complex systems, but also shows its unique value in the analysis of biological, social science, economic and technological networks. Multi-layer networks pay special attention to how the interactions between different types of network layers affect the dynamics and function of the overall network. By simulating the interactions between these layers, researchers can better explain how systems work together at multiple levels.
[0004] The introduction of evolutionary game theory provides a powerful tool for understanding complex decision-making processes and competitive behaviors. It not only improves the understanding of network behavior, but also promotes the development of new algorithms, which are now being used to optimize network design, enhance system robustness, and predict changes in socioeconomic patterns.
[0005] However, there are few studies that combine multi-layer networks and evolutionary games at the same time. The combined studies often only consider the conclusions of simple migration on a single-layer complex network or ignore the difference in strategies and blindly accumulate the benefits on multiple layers. These actions are somewhat different from reality. These gaps make the current algorithms too dependent on external parameter transmission in improving the group cooperation rate and perform generally, thus increasing the requirements for subsequent operations and causing trouble for subsequent experiments. Summary of the invention
[0006] In view of this, the purpose of the present invention is to propose a dynamic collaboration method of intelligent robots based on a multi-layer network and a difference accumulation strategy, which continuously optimizes its own behavior strategy through real-time feedback and continuous evolutionary algorithms to adapt to the ever-changing working environment and task requirements.
[0007] The technical means adopted by the present invention are as follows:
[0008] A method for dynamic collaboration of intelligent robots based on a multi-layer network and a difference accumulation strategy comprises the following steps:
[0009] S1. Obtain the current network characteristics of the intelligent robot, and establish a multi-layer network model using the GM algorithm based on the current network characteristics;
[0010] S2. Select the game model, game matrix parameters and game mixing ratio according to the current network characteristics to obtain the game matrix; initialize the robot strategy;
[0011] S3. All intelligent robots interact with other intelligent robots within their own interaction range in all network layers of the multi-layer network model;
[0012] S4. Calculate the game benefits and fitness according to the current game matrix, and use the difference accumulation strategy to correct the fitness;
[0013] S5, update the robot strategy of S2 according to the fitness corrected by S4;
[0014] S6. Determine whether the system has reached a steady state. If so, end the game. If not, return to S3.
[0015] Furthermore, S1 specifically includes the following steps:
[0016] S11. Use the Barabási-Albert Model to generate n BA networks containing N nodes, so that the number and position of nodes and robots on each layer of the BA network correspond one to one, and initialize the interaction radius parameter r of each node in the BA network. i ;According to the degree k of each node and the degree distribution of the network obtained in the BA network;
[0017] S12. Calculate the connection probability of nodes between any adjacent BA networks, and connect adjacent BA networks according to the connection probability, where the connection probability between node i and node j is calculated as follows:
[0018]
[0019] The parameter T∈[0,1) controls the degree of network aggregation. When T=0, the network aggregation coefficient is the largest. The parameter R~lnN; x ijUsed to represent the hyperbolic distance between i and j, the calculation formula is as follows:
[0020] x ij =r i +r j +2lnsin(Δθ ij / 2)
[0021] Δθ ij =k i k j / (L i L i )
[0022] Among them, r i Abstractly represents the interaction radius of node i, r i The smaller the value of , the larger the interaction radius of the node; ij Used to indicate the degree of similarity between nodes, k i represents the degree of i, L i Indicates the total number of connections in the network where i is located;
[0023] S13, use Xalvi-Brunet and Sokolov algorithm to adjust the multi-layer network connected in S12; randomly select two edges ab and cd with the same properties, exchange the nodes to obtain ac and bd;
[0024] S14, re-count the characteristics of the changed multi-layer network. If the new characteristics are closer to the current network characteristics of S1 than before, then keep the change, otherwise discard it;
[0025] S15, determine whether the network model generated by S14 is equal to or close to the current network characteristics of S1; if so, end the process and obtain a multi-layer network model; if not, return to S13 to continue adjustment.
[0026] Furthermore, in S2, the game matrix is assigned values according to actual data, and the game model includes a prisoner's dilemma matrix and a public goods game, and the prisoner's dilemma matrix and the public goods game only occur between individuals at the same level;
[0027] S2 specifically includes the following steps:
[0028] S21, initialize the parameters of the prisoner's dilemma game; the parameters of the prisoner's dilemma game include the specific values of reward R, punishment P, temptation T and double loss S; during the initialization process, make T>R>P>S and T+S<2R;
[0029] S22, initializing public goods game parameters; public goods game parameters include contribution value c, benefit b and growth coefficient r;
[0030] S23. Set different game proportions p so that the network layer with a proportion of p adopts the prisoner's dilemma game model in the game.
[0031] Furthermore, S3 specifically includes the following steps:
[0032] S31. Initialize the initial strategies of the points on all levels. Each point has the same probability of choosing to cooperate with C or betray D. The strategies of the same robot in different network layers are independent of each other.
[0033] S32, initialize the tendency of using the difference accumulation strategy for points on all levels to p∈(0,1], and the parameter distribution follows the normal distribution;
[0034] S33, all points on the level play games with other points within their own interaction radius; if the prisoner's game is adopted in this level, the payoff matrix is:
[0035]
[0036]
[0037] The node determines its own income π based on its own strategy and the strategy of the object it interacts with. ijl , i and j represent the two individuals interacting, and l represents the network level. If the layer is a public goods game, the profit calculation formula is as follows: i and j represent that node i participates in the game centered on j, and l represents the network level.
[0038]
[0039] Furthermore, S4 specifically includes the following steps:
[0040] S41. Compute the benefits of nodes on layer m:
[0041]
[0042] S42. Obtain fitness based on the game income level income and level weight, the formula is as follows:
[0043]
[0044] S43. Determine whether to modify the fitness according to the trend of the node adopting the difference accumulation strategy. If so, then:
[0045]
[0046] Among them, s represents the strategy adopted by the central node, The representatives and the central nodes adopt different strategies;
[0047] Otherwise, the fitness remains unchanged.
[0048] Furthermore, S5 specifically includes the following steps:
[0049] S51: Node i randomly selects an object node j within its interaction range and obtains its fitness U;
[0050] S52: Calculate the strategy imitation probability. The probability that node i on layer l imitates node j is:
[0051]
[0052] Among them, K quantifies the irrationality of the node in the game and is set to 0.1;
[0053] S53: Imitation probability determines whether the update operation occurs. If it occurs, the robot strategy is updated, otherwise it is not updated;
[0054] S54: Perform S51 to S53 on all nodes in the network.
[0055] Furthermore, S6 specifically includes the following steps:
[0056] S61: Obtaining the single-layer cooperation rate c l ;
[0057] S62: Get system cooperation rate:
[0058]
[0059] S63: If It indicates that the system has reached a steady state, c is the current cooperation rate, is the average of the cooperation rates under the three evolutionary steps, execute S65; otherwise, execute S64;
[0060] S64: Determine whether the number of evolution rounds has reached the upper limit, if so, proceed to step S64, otherwise, return to step S3;
[0061] S65: Complete the evolution, record and persist all current individual strategies and system cooperation rates, and perform strategy migration.
[0062] The present invention also provides a storage medium, which includes a stored program, wherein when the program is run, any of the above-mentioned intelligent robot dynamic collaboration methods based on a multi-layer network and a difference accumulation strategy is executed.
[0063] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes any one of the above-mentioned intelligent robot dynamic collaboration methods based on a multi-layer network and a difference accumulation strategy through the computer program.
[0064] Compared with the prior art, the present invention has the following advantages:
[0065] The present invention adopts a new multi-layer network modeling method, and performs multi-layer network modeling according to the collected data, so that the result is closer to the actual situation;
[0066] The present invention adopts mixed game to describe the game between individuals, which is more accurate and has wider transferability;
[0067] The present invention adopts a difference accumulation strategy to assist the individual updating strategy, which improves the cooperation rate and makes the system faster.
[0068] The present invention adds a new steady-state judgment mechanism to make the judgment more consistent with the characteristics of complex systems, improve the accuracy and save experimental time.
[0069] When the present invention is applied to the field of port logistics, the intelligent robot using the method considers various information and continuously optimizes its behavior strategy through real-time feedback and continuous evolutionary algorithms to adapt to the ever-changing working environment and task requirements.
[0070] When the present invention is applied to the field of drone assembly, the multi-layer network of the present invention can more accurately describe the actual situation, and real-time feedback and continuous evolution can enable the drone to make accurate decisions immediately, thereby achieving the goal faster. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0072] Figure 1 It is the overall flow chart of the present invention;
[0073] Figure 2 Constructing a multi-layer network model for the present invention;
[0074] Figure 3 It is a flow chart of the evolutionary game stage of the present invention. DETAILED DESCRIPTION
[0075] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0076] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0077] like Figure 1-3 As shown, the present invention provides a method for dynamic collaboration of intelligent robots based on a multi-layer network and a difference accumulation strategy. The scheme can be mainly divided into two main modules, a network model building module and an evolutionary game module;
[0078] like Figure 2 As shown, the process of S1 building the network model is as follows:
[0079] The specific method used in S1 is as follows:
[0080] S11. Use the Barabási-Albert Model to generate n BA networks containing N nodes, ensure that the nodes on each layer correspond to the robots one by one, and initialize the interaction radius parameter r of each node. i . According to the generated network, the degree k of each point and the degree distribution of the network are obtained.
[0081] S12, according to calculating the connection probability of nodes between any adjacent networks, the connection probability between node i and node j is calculated as follows:
[0082]
[0083] Among them, the parameter t∈[0,1) controls the degree of network aggregation. When T=0, the network aggregation coefficient is the largest; the parameter R~lnN. ij Used to represent the hyperbolic distance between i and j, the calculation formula is as follows:
[0084] x ij =r i +r j +2lnsin(Δθ ij / 2)
[0085] Δθ ij =k i k j / (L i L i )
[0086] r i Abstractly represents the interaction radius of node i, r i The smaller the value of , the larger the interaction radius of the node; ij Used to indicate the degree of similarity between nodes, k i represents the degree of i, L i Indicates the total number of connections in the network where i is located.
[0087] S13. Use the Xalvi-Brunet and Sokolov algorithm to adjust the connected multi-layer network; randomly select two edges ab and cd with the same properties (cross-layer or same-layer), and exchange the nodes to obtain ac and bd.
[0088] S14, re-counting the characteristics of the multi-layer network, if the change is close to the standard, it is retained, otherwise it is discarded;
[0089] S15: Determine whether the generated network model meets the standards. If so, end the S1 process and obtain the model; if not, return to step S13 to continue adjustment.
[0090] The specific method used in S2 is as follows:
[0091] The game matrix is assigned values according to actual data. The game models used include the prisoner's dilemma matrix and the public goods game. These interactions only occur between individuals at the same level and do not cross levels. By implementing different game models, the present invention shows how individuals interact with different objects in various aspects in social life. These models can effectively simulate and predict the behavior patterns of individuals in different social interaction scenarios.
[0092] S21: Initialize the parameters of the Prisoner's Dilemma game; including but not limited to the specific values of reward R, punishment P, temptation T, and double loss S. These parameters are key factors in game theory for defining various possible strategy results, and their settings directly affect the accuracy and practicality of the simulation results. During initialization, ensure that t>R>P>S and T+S<2R.
[0093] S23: Initialize the public goods game parameters, including contribution value c, benefit b and growth coefficient r;
[0094] S24: Set different game proportions p so that the network layer with a proportion of p adopts the prisoner's dilemma game model in the game.
[0095] like Figure 3 As shown in the figure, the evolutionary game module mainly includes:
[0096] S3. All intelligent robots interact with other intelligent robots within their interaction range in all network layers;
[0097] S4. Calculate the game benefits and fitness according to the current game matrix, and use the difference accumulation strategy to correct the fitness;
[0098] S5. Update your strategy according to fitness;
[0099] S3 specifically adopts the following methods:
[0100] S31. Initialize the initial strategies of the points on all levels. Each point has the same probability of choosing cooperation (C) or betrayal (D). The strategies of the same robot in different network layers are independent of each other.
[0101] S32. Initialize the tendency of using the difference accumulation strategy for points on all levels to be p∈(0, 1], and the parameter distribution obeys the normal distribution.
[0102] S33, all points on the level play games with the other points within their own interaction radius; if the prisoner's game is adopted in this level, the payoff matrix is
[0103] C D C R S D T P
[0104] The node determines its own income π based on its own strategy and the strategy of the object it interacts with. ijl , i and j represent the two individuals interacting, and l represents the network level; if the layer is a public goods game, the profit calculation formula is as follows, i and j represent that node i participates in the game centered on j, and l represents the network level.
[0105]
[0106] S4 specifically adopts the following method:
[0107] S41. The calculated benefit of the node on layer m is
[0108] S42. Obtain fitness based on the game income level income and level weight. The formula is as follows
[0109]
[0110] S43, judging whether to modify the fitness according to the trend of the node adopting the difference accumulation strategy, if adopted, Among them, s represents the strategy adopted by the central node, The representative and the central node adopt different strategies; otherwise, the fitness remains unchanged.
[0111] S5 specifically adopts the following method:
[0112] S51, node i randomly selects an object node j within its interaction range and obtains its fitness U;
[0113] S52: Calculate the strategy imitation probability. The probability that node i on layer l imitates node j is
[0114]
[0115] K quantifies the irrationality of the node in the game and is set to 0.1 here.
[0116] S53: Determine and perform a change operation.
[0117] S54: Perform the above operation on all nodes in the network.
[0118] S6 specifically adopts the following method:
[0119] S61. Obtaining single-layer cooperation rate c l
[0120] S62. Obtain system cooperation rate
[0121]
[0122] S63, if It indicates that the system has reached a steady state, c is the current cooperation rate, is the average of the cooperation rates under the three evolutionary steps, executing S65; otherwise, proceeding to the next step of judgment;
[0123] S64, determine whether the number of evolution rounds has reached the upper limit, if so, enter, otherwise, return to S3;
[0124] S65. Complete the evolution, record and persist all current individual strategies and system cooperation rates, and perform strategy migration.
[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A dynamic collaboration method of intelligent robots based on multi-layer networks and difference accumulation strategy, characterized in that: The steps include: S1. Obtain the current network characteristics of the intelligent robot, and establish a multi-layer network model using the GM algorithm based on the current network characteristics; S2. Select the game model, game matrix parameters and game mixing ratio according to the current network characteristics to obtain the game matrix; initialize the robot strategy; S3. All intelligent robots interact with other intelligent robots within their own interaction range in all network layers of the multi-layer network model; S4. Calculate the game benefits and fitness according to the current game matrix, and use the difference accumulation strategy to correct the fitness; S5, update the robot strategy of S2 according to the fitness corrected by S4; S6. Determine whether the system has reached a steady state. If so, end the game. If not, return to S3.
2. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: S1 specifically includes the following steps: S11. Use the Barabási-Albert Model to generate n BA networks containing N nodes, so that the number and position of nodes and robots on each layer of the BA network correspond one to one, and initialize the interaction radius parameter r of each node in the BA network. i ;According to the degree k of each node and the degree distribution of the network obtained in the BA network; S12. Calculate the connection probability of nodes between any adjacent BA networks, and connect adjacent BA networks according to the connection probability, where the connection probability between node i and node j is calculated as follows: Among them, the parameter T∈[0,1) controls the degree of network aggregation. When T=0, the network aggregation coefficient is the largest; the parameter R~lnN; x ij Used to represent the hyperbolic distance between i and j, the calculation formula is as follows: x ij =r i +r j +2lnsin(Δθ ij / 2) Dth ij =k i k j / (L i L i ) Among them, r i Abstractly represents the interaction radius of node i, r i The smaller the value of , the larger the interaction radius of the node; ij Used to indicate the degree of similarity between nodes, k i represents the degree of i, L i Indicates the total number of connections in the network where i is located; S13, use Xalvi-Brunet and Sokolov algorithm to adjust the multi-layer network connected in S12; randomly select two edges ab and cd with the same properties, exchange the nodes to obtain ac and bd; S14, re-count the characteristics of the changed multi-layer network. If the new characteristics are closer to the current network characteristics of S1 than before, then keep the change, otherwise discard it; S15, determine whether the network model generated by S14 is equal to or close to the current network characteristics of S1; if so, end the process and obtain a multi-layer network model; if not, return to S13 to continue adjustment.
3. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: In S2, the game matrix is assigned values according to actual data. The game model includes a prisoner's dilemma matrix and a public goods game. The prisoner's dilemma matrix and the public goods game only occur between individuals at the same level. S2 specifically includes the following steps: S21, initialize the parameters of the prisoner's dilemma game; the parameters of the prisoner's dilemma game include the specific values of reward R, punishment P, temptation T and double loss S; during the initialization process, make T>R>P>S and T+S<2R; S22, initializing public goods game parameters; public goods game parameters include contribution value c, benefit b and growth coefficient r; S23. Set different game proportions p so that the network layer with a proportion of p adopts the prisoner's dilemma game model in the game.
4. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: S3 specifically includes the following steps: S31. Initialize the initial strategies of the points on all levels. Each point has the same probability of choosing to cooperate with C or betray D. The strategies of the same robot in different network layers are independent of each other. S32. Initialize the tendency of using the difference accumulation strategy for points on all levels to p∈(0,1], and the parameter distribution follows the normal distribution; S33, all points on the level play games with other points within their own interaction radius; if the prisoner's game is adopted in this level, the payoff matrix is: The node determines its own income π based on its own strategy and the strategy of the object it interacts with. ijl , i and j represent the two individuals interacting, and l represents the network level. If the layer is a public goods game, the profit calculation formula is as follows: i and j represent that node i participates in the game centered on j, and l represents the network level.
5. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: S4 specifically includes the following steps: S41. Compute the benefits of nodes on layer m: S42. Obtain fitness based on the game income level income and level weight, the formula is as follows: S43. Determine whether to modify the fitness according to the trend of the node adopting the difference accumulation strategy. If so, then: Among them, s represents the strategy adopted by the central node, The representatives and the central nodes adopt different strategies; Otherwise, the fitness remains unchanged.
6. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: S5 specifically includes the following steps: S51: Node i randomly selects an object node j within its interaction range and obtains its fitness U; S52: Calculate the strategy imitation probability. The probability that node i on layer l imitates node j is: Among them, K quantifies the irrationality of the node in the game and is set to 0.1; S53: Imitation probability determines whether the update operation occurs. If it occurs, the robot strategy is updated, otherwise it is not updated; S54: Perform S51 to S53 on all nodes in the network.
7. The intelligent robot dynamic collaboration method based on multi-layer network and difference accumulation strategy according to claim 1 is characterized in that: S6 specifically includes the following steps: S61: Obtaining the single-layer cooperation rate c l ; S62: Get system cooperation rate: S63: If It indicates that the system has reached a steady state, c is the current cooperation rate, is the mean value of the cooperation rate under three evolutionary steps, and executes S65; Otherwise, execute S64; S64: Determine whether the number of evolution rounds has reached the upper limit, if so, proceed to step S64, otherwise, return to step S3; S65: Complete the evolution, record and persist all current individual strategies and system cooperation rates, and perform strategy migration.
8. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the intelligent robot dynamic collaboration method based on a multi-layer network and a difference accumulation strategy as described in any one of claims 1 to 7 is executed.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The processor executes the intelligent robot dynamic collaboration method based on a multi-layer network and a difference accumulation strategy as described in any one of claims 1 to 7 through the computer program.
Citation Information
Patent Citations
Multi-robot cooperation system and method based on third-order reputation mechanism and mixed game
CN117875086A
Robot dynamic cooperation method based on reputation satisfaction and evolutionary game theory
CN117984338A
Data service-oriented adaptive intrusion response game method and system thereof
WO2021180017A1
